The Clipboard Was Better

I spent last week on vacation with friends and family in Lake George, New York. It was our 6^th year going to the same place.

It’s a YMCA family camp kind of place where you rent a cabin, and then they have activities all week. Kayaking, ropes course, archery, etc.

It was all very simple and very quaint. In the morning, we’d walk into the main lobby, look at the day's activities, and sign up for what we wanted on a clipboard.

Between the three families, we have 10 kids and 16 people total. So I’d just sign up and write “16 people.” It worked great.

This year, they created a new ‘digital experience.’

To sign up for an activity, you’d have to go in 16 different times, once for each person, and sign them up individually. It took 30 minutes per activity. So, 90 minutes of sign-up time to do three activities for the day.

I asked the front desk if she could just make a manual note in the system that we’d have 16 people attending for a specific time slot to avoid the bureaucracy.

She told me she couldn’t, that the ONLY way was for us to manually sign up each person for each activity that we wanted to do.

So, for the first time, we didn’t do the activities. Instead, we rented a boat for a day, went to a ropes course off campus, and went to a theme park. All of that was easier than dealing with the administrative friction of their ‘digital experience’.

The FRICTION of signing up for the activities meant we just didn’t do the activities. Which is the opposite of what the YMCA wanted. Perhaps the digital sign-up made it easier for the staff, but it sure didn’t improve the guest experience.

Earlier this summer, I was doing a backyard project with my kids and ordered supplies from Home Depot. 30 bags of concrete; some tools; grass seed. Basic stuff.

I didn’t want to pick up 30 bags of concrete, so I placed the online order.

What a mistake.

It took me 3 hours on the phone to actually get the materials delivered because they were coming from multiple stores (as if that were my problem to deal with).

But even that wasn’t simple.

First, I called the online number, and they told me I had to call each store individually to coordinate delivery.

Then I called the store. And they told me I had to call the online number since I ordered through the app.

After about an hour, I asked the rep if he could just get it all figured out and call me back.

He said he couldn’t because if he wasn’t on the phone with me, he wouldn’t be able to access my account.

So I had to sit on the phone with him for another 2 hours while he called the stores and coordinated everything. So Home Depot hijacked my phone. I couldn’t say or do anything during the two hours because I had to be on the phone. If I hung up, another call would come in, and he wouldn’t be able to work on my issue. And guess what I would have to go through again?

I’m not done yet.

When they finally did the delivery, two things weren’t included. So I asked the driver. He told me to call the store. So I called the store. And guess who they blamed? The driver. So I decided to just cut to the chase and ask when they could deliver them. They told me they couldn’t because the system showed my delivery as completed.

In the end, I just gave up, went to the store myself, and bought the two items.

Which I probably should have done the first time around.

Every business has something like these two examples.

But before getting into how this plays out in the mortgage industry, I’d like to start by talking about one of the biggest mysteries in human history.

I have a friend from college who was desperate to get a girlfriend and get married. He called me one day while I was driving and told me he was going to spend $30K on a ‘dating coach.’

I suggested that before he spends the money, maybe he should talk to my wife Sarah and a few of her friends about ‘what women want.’

Not knowing he was on speakerphone with my wife in the car, he said: “That’s pointless. Women don’t know what they want.”

My wife then chimed in: “Thinking that is why you’re single and the $30k won’t get you anywhere.”

I’ve been with my wife for over 20 years. I still find it productive to regularly ask her what she wants. It’s a lot better than guessing or trying to read her mind. Or getting advice from the online manosphere.

So what do borrowers want? And how do you find out?

To answer that question, we launched a brand promise eight years ago called The Princeton Promise. The idea behind it is simple:

We commit to delivering an effortless experience. If you are unhappy for any reason, we’ll give you $1,000.”

Anytime someone claims the Princeton Promise, we give them the $1,000 and speak with them.

It’s the best money I’ve ever spent.

On vacation, I just wanted to do the activities without the friction of signing up for them.

At Home Depot, I just wanted to get the supplies for my backyard project.

Borrowers want the same thing.

They want to get a mortgage with a fair rate and costs from someone who understands their unique situation and operates with fiduciary responsibility, and to do so with as little friction as possible.

The Princeton Promise is my friction detector.

Sounds great. But as it turns out, reducing friction is the hardest part of the mortgage industry.

I often hear people blame the borrower for issues with a loan. Perhaps, like 2% of the time.

But 98% of the time (my own stat), borrowers aren’t the bottleneck.

Friction is.

Here is what some of that friction can look like:

• Asking for a document they already gave us

• Clearing a condition via email instead of in the POS

• A condition with mortgage language or no reason/explanation

• A borrower has to follow up because they don’t understand

• Waiting, with no idea where they stand

• Telling the originator, the LOA, and the processor the same thing because it’s stuck in each person’s head.

LenderLogix released an interesting report this week covering 3,000 mortgage applications:

• 88.5% of borrowers complete their app in under 30 minutes

• 72.77% of borrowers upload their documents within 24 hours of completing the App. 53.4% do so within 1 hour of completing the app.

And then the borrowers wait.

Conditional approval to CTC is the slowest, most manual part of the process and the highest source of borrower effort.

Two years ago, we built an orchestration and execution layer. In the beginning, it was just sort of like: “How can we better organize our loan data to know what we should be prioritizing and see it all in one spot?”

Now it’s our workflow engine and embedded AI.

Last month, we launched a workflow called “24 to Close”

The idea was to reduce both the effort required of our borrowers and our team.

The loan comes out of underwriting with a Conditional Approval. The software grabs the conditions, rewrites them with AI, allows the processor to edit the conditions, and excludes the ones they don’t want to send.

Everything gets formatted beautifully and sent to the borrower. Standardized across all processors, borrowers, and originators. When the borrower clicks the condition to upload a doc, it takes them right to the condition folder in the POS. The condition folder in the POS has AI that reads the document.

If the system has high confidence that the document fulfills the condition, it can be pushed right into underwriting.

Very soon, AI will be able to read the documents and find a problem, say, the wrong month on the bank statement or a missing page. It can chat in real time with the borrower as they are uploading documents.

Compare that with what happens today.

The borrower uploading the wrong thing, the processor noticing it a few days later, submitting it to UW, and then the UW catching it. The processor has to go back to the borrower to get the update.

All of that is Friction.

But the friction can now be removed in real time.

But you have to be careful.

Notice who this system was built for. Yes, it makes things easier on our side. But that just so happened to be what needed to happen in order to make the experience for the buyer.

The minute you start building systems or using AI to only remove your own friction, you're going to run into a lot of problems. The first one being, by making things easier for yourself, you will, like the camp on Lake George or Home Depot, make this harder for your customers.

And while that may not be a big deal for a little while. The minute someone else comes along with a focus on eliminating friction for the buyer, you will be in big trouble.

I know this firsthand because I was a little stressed while we were rolling out “24 to Close,” as many of our originators and processors are very proud of how they handle the conditional approval-to-CTC process. And I don’t blame them. It’s their own manually designed process that allowed them to outperform.

First off, the resistance comes from good intentions. They want to work harder to make things easier for the buyer. Secondly, I’ve learned not to force change on people. It just doesn’t work. So what I did instead was build alternative ways of doing things and let them bounce back and forth. Because if what we built is better than the way they were doing it, they will inevitably gravitate to the new way. Because, like buyers, they also want things to be easier. Once you win over the most vocal, stubborn people, we roll it out to everyone else.

And that is exactly what happened.

Here’s what two of our most opinionated top producers emailed me. Originators who, previous to using the workflow, were very attached to the way they had their processors doing things.

“So glad you’re giving this 24-to-close the attention it needs. We’re hearing “that was easy” from clients.”

“It’s great to all be on the same page for processing and chasing documentation as a unit.”

As Benjamin Franklin once said, “There is no point in denying human nature.”

And as Jeff Bezos once advised: “Design, build, and deliver what your customers want at a price you can turn a profit.”

Your business, and my business, is actually that simple. Figure out what your customer wants, and give it to them at a price you can turn a profit.

In 2018, we had 3 people and no loans. I had to develop a differentiated value proposition to gain market share from incumbents. After reading the book “The Effortless Experience” by Matthew Dixon. I realized that the key was not just to reduce friction, but to reduce friction for the right people.

In the book, they describe studying 97,000 customer service interactions to prove that exceeding expectations builds loyalty. It doesn’t. What destroys loyalty is making the customer work, repeat themselves, chase an answer, use multiple systems, talk to multiple people, and wonder what happens next.

According to their data, 96% of high-effort customers become disloyal. Only 9% of low-effort customers turn disloyal.

Loyalty isn’t won by doing extra.

It’s lost by making them do extra.

So instead of getting your borrowers to say that you ‘exceeded their expectations,’ your goal really ought to be trying to get customers to say, ‘you made that easy.’

Back in 2018, the only tools we really had at our disposal to reduce effort were training our team to ‘do better’ and ‘work harder’.

As I have already demonstrated, AI changes all that.

AI has enabled us to extract data from documents accurately and efficiently.

My next goal is to use it to lower software development costs by 90%. so that what used to cost $250k to build now costs us $25k. We can FAFO now. And it’s fun. And it’s working.

You can measure the degree of differentiation among value propositions in the mortgage industry by looking at their Cost Per Funded Loan. Cost per loan is a direct measure of the machine's effectiveness.

From 2008 through 2009, the gap between the bottom 20% and top 20% cost performers was only $941 per loan. Very little cost differentiation between the best and worst.

From 2020 to 2025, the top 20% cost performers had an expense of $10,074, while the bottom 20% averaged $12,603.

The spread between efficient and inefficient lenders has nearly tripled, from $941 to $2,626.

My argument is simple.

Be at a company where you can:

1. Offer competitive rates

2. Earn 40-50% earnings margin (W2 wages / Gross Revenue on your loans)

3. Deliver low friction to your borrowers, not because you or your processor are amazing, but because the system is designed around reducing effort for your borrowers.

Rich Weidell
CEO, Princeton Mortgage